Last Changes
Helmut Weick, 2nd Feb 2020

Remaining problems:

"S L" and "A C" work only in CM1, "R M" only in CM1 or CM2.
In combined multipoles in C M 1 the quadrupole strength must not be zero or very close to zero.
Maybe some error in third order coefficients in inhomogenous dipoles in CM1.
In the plot of electrostatic dipoles some lines are missing.
Electrostatic acceleration column (A C) has errors in higher order chromatic terms.
system matrix output of path length dispersion [L,D] in CM1 with option (O C * * L *) not correct, use CM2 or CM3,
but no error in output of all single matrices with P A B E (important for MOCADI).

Pitfalls:
The system matrix (overall transfer matrix of the full ion-optical system) always contains an object lens which is used to describe a not upright phase space ellipse, even when making the input in mode TX, TY.

Internally GICOSY always calculates in coordinates energy, time and tangens of angle in symplectic coordinates. That means any direct assignment of matrix coefficients - for example to specify the initial dispersion - has to be made in these coordinates. However, for all output the coordinate system can be changed by the command OC.

The variable L in calculation mode CM2 or CM3 means path length difference, but in CM1 it is the time difference divided by the reference velocity.
Only for output of single element matrices (P A) the variable L really is the path length difference also in CM1.

2020 Added missing lines for Y dispersion in DB/DS modes 1+3 -> V2.8

2016 Negative radius as input for dipole deflection direction now also works for CM3 -> V2.7

2016 Added MIRKO style plot with y envelop in upper half and y envelope in lower half of plot, with dispersion on x on top of both. -> V2.6

2009 Corrected variable assumed static which caused problems in phase space plots on some machines -> V2.5

2008 Corrected error in shift of reference trajectory in the floor coordinates for sector fields with rho<0 -> V2.4

2007 Corrected two bugs -> V2.3
The initial Twiss parameter alpha was to large by a factor of two in some calculations
Dispersive matrix coefficients in mode momentum were wrong in expression plots.
Corrections in conversion to postscript (meta2ps).

2006 Enhanced size of static arrays for more optical elements.

2005 Renamed to GICOSY.

The oldest reference to GICO is from November 1990, but the name GICO was already protected several months earlier.
As a consequence also the names of input/output files were changed.

1998 Added functions already existing in GSI version of GICO

P A, options B E to write all element matrices to file for MOCADI.
SM and IM, option to write/read ASCII files
K E, a wedge degrader defined 1st. order transfer matrix

1995 GICO found in Giessen installed on UNIX of DEC alpha machines.

Added cards:
AC, acceleration column (has some higher order chromatic bugs)
RM, rotation matrix for skewed elements and beam bending out of plane
DB, more options (3+4) for 3D phase space
SL, solenoid based on paper by Sakurai with correction by M. Yavor.

The program is based on COSY 5.0 by Martin Berz, many people have added to it at the II. Physikalisches Institut of the University of Giessen.

This page was last updated by Dr. Helmut Weick, 25^th May 2018, contact h.weick(at)gsi.de, Imprint (Impressum), Privacy Policy (Datenschutzerklärung)